it's odd I haven't noticed it before. In other news, I've learnt how to
do image captions.
In virus news, James has called the peak, so it's all downhill from here. ATTP has a post I like on models, which almost but doesn't quite say what I've been thinking, which is that the epidemic modelling isn't really ready for the big time. Which brings me back to ...and global warming. Anything you do, and don't test, will be wrong.
Somewhat related, Interventions are likely to be fat-handed with incentive research as a case from TF. Which starts off with some cognitive-bias research which finds that even large incentives don't remove the bias; but then segues into the nice thought that this is only true if you leave the pool of those solving the problem untouched; but large incentives would of course attract those able to better solve the problem. Which is another way of expressing the advantage of the free market.
Contact tracing via smartphone is very much in vogue as an idea, so I'll point you at LBT's Contact Tracing in the Real World. If Google does it, it will be competent; but if (as the news appears to suggest) the NHS does it... we'll be perfectly safe from it for a couple of years. Incidentally, determining range from BT (certainly from BREDR; might be easier from LE) has a long history of being Difficult.
Refs
* Will Fermi and Dirac Save Us All? Probably Not.
* A failure, but not of prediction - SlateStarCodex
* How to Win Friends and Influence People Book Club, Part 4 by Bryan Caplan
* Why do human beings keep getting diseases from bats?
* Markets, Morality, and Mises by G. Patrick Lynch at EconLib
3 comments:
"The Logic of Chance" by Eugene Koonin
For a long and difficult read.
William's wasp nest displays a weird consilience with orbital images of the atmosphere of Jupiter.
It may be time for SETI to call the exterminators, and launch a search for pyrethrin molecules in the Kuiper belt.
All models are wrong.
Some models are useful, with care.
Post a Comment